Debian Developers - When does a conflict become outdated?

This is Interesting: Free IT Magazines  
Home > Archive > Debian Developers > April 2004 > When does a conflict become outdated?





You are viewing an archived Text-only version of the thread. To view this thread in it's original format and/or if you want to reply to this thread please [click here]

Author When does a conflict become outdated?
Morten Brix Pedersen

2004-03-31, 10:40 am

Hi,

Many old packages in Debian, have conflicts for specific version of
packages that are pre-potato. Would it then be OK to submit a minor bug
and tell the maintainer that the conflict could be removed?

For example, the debconf package has Conflicts: menu (<= 2.1.3-1). But
even potato has menu 2.1.5-10.1, so the conflict is not of much use
anymore.

- Morten.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Zefram

2004-03-31, 10:40 am

Morten Brix Pedersen wrote:
>For example, the debconf package has Conflicts: menu (<= 2.1.3-1). But
>even potato has menu 2.1.5-10.1, so the conflict is not of much use
>anymore.


Looks perfectly useful to me. It is most definitely not a bug.

-zefram


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Steve Greenland

2004-03-31, 10:40 am

On 31-Mar-04, 08:16 (CST), Morten Brix Pedersen <morten@mbrix.dk> wrote:
> Hi,
>
> Many old packages in Debian, have conflicts for specific version of
> packages that are pre-potato. Would it then be OK to submit a minor bug
> and tell the maintainer that the conflict could be removed?


Why? What harm is it doing? And don't say "bloat": the space taken by
all the outdated conflicts combined could be easily compensated for
by removing, say one IRC client. Or one editor. Or one module from
xscreensaver.

And if you're going chase down obsolete conflicts, then you should also
chase down all the parts of {pre,post}{inst,rm} scripts that handle
upgrades from ancient versions, etc.

Steve

--
Steve Greenland
The irony is that Bill Gates claims to be making a stable operating
system and Linus Torvalds claims to be trying to take over the
world. -- seen on the net


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Matt Zimmerman

2004-03-31, 1:37 pm

On Wed, Mar 31, 2004 at 08:27:15AM -0600, Steve Greenland wrote:

> On 31-Mar-04, 08:16 (CST), Morten Brix Pedersen <morten@mbrix.dk> wrote:
>
> Why? What harm is it doing? And don't say "bloat": the space taken by
> all the outdated conflicts combined could be easily compensated for
> by removing, say one IRC client. Or one editor. Or one module from
> xscreensaver.


It degrades the readability of the dependency relationships, and distracts
the maintainer from the important ones.

--
- mdz


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Steven Augart

2004-03-31, 3:38 pm

Matt Zimmerman wrote:
> On Wed, Mar 31, 2004 at 08:27:15AM -0600, Steve Greenland wrote:
[...][color=darkred]
[color=darkred]
[...][color=darkred]
> It degrades the readability of the dependency relationships, and distracts
> the maintainer from the important ones.


Precisely. The core problem in software engineering is control of
complexity. If, as may be the case here, we can reduce complexity
without sacrificing other useful qualities, it is good to do that.

--
Steven Augart

Jikes RVM, a free, open source, Virtual Machine:
http://oss.software.ibm.com/jikesrvm


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Sebastian Ley

2004-03-31, 6:35 pm

Am Mi, den 31.03.2004 schrieb Matt Zimmerman um 20:29:

>
> It degrades the readability of the dependency relationships, and distracts
> the maintainer from the important ones.


I do not remember offhand, but IIRC we _ensure_ upgrade paths only from
one version to a subsequent one. So anything beyond this is a bonus for
the maintainer to decide. I agree that removing cruft from the
dependencies improves readability and should thus be considered by
maintainers.

Sebastian

--
PGP-Key: http://www.mmweg.rwth-aachen.de/~se....ley/public.key
Fingerprint: A46A 753F AEDC 2C01 BE6E F6DB 97E0 3309 9FD6 E3E6



--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Adrian 'Dagurashibanipal' von Bidder

2004-04-01, 1:34 am

Marc Haber

2004-04-01, 2:34 am

On Wed, 31 Mar 2004 16:16:02 +0200, Morten Brix Pedersen
<morten@mbrix.dk> wrote:
>Many old packages in Debian, have conflicts for specific version of
>packages that are pre-potato. Would it then be OK to submit a minor bug
>and tell the maintainer that the conflict could be removed?
>
>For example, the debconf package has Conflicts: menu (<=3D 2.1.3-1). But
>even potato has menu 2.1.5-10.1, so the conflict is not of much use
>anymore.


I tend to keep these conflicts in to cater for the stupidity of our
users. To give a reason: exim4's maintainer scripts create a user
account. This account creation fails if the version of passwd is in a
certain range due to a bug in passwd. The buggy versions are newer
than woody, but older than testing, so any Debian system running a
current version of any distribution is fine.

However, there have been gazillions of bug reports against exim
because people keep running outdated versions of unstable. We hat to
add a versioned conflicts just to keep the number of bug reports down,
while that conflict is technically not necessary any more. And no,
that conflicts is bound to stay for a while. Won't be removed any time
soon.

Greetings
Marc

--=20
-------------------------------------- !! No courtesy copies, please !! =
-----
Marc Haber | " Questions are the | Mailadresse im =
Header
Karlsruhe, Germany | Beginning of Wisdom " | Fon: *49 721 966 32=
15
Nordisch by Nature | Lt. Worf, TNG "Rightful Heir" | Fax: *49 721 966 31=
29
Julian Mehnle

2004-04-01, 3:37 am

Steve Greenland wrote:
> Morten Brix Pedersen <morten@mbrix.dk> wrote:
>
> Why? What harm is it doing? And don't say "bloat" [...]


No, it's about complexity. Complexity that is not needed any more.

> And if you're going chase down obsolete conflicts, then you should also
> chase down all the parts of {pre,post}{inst,rm} scripts that handle
> upgrades from ancient versions, etc.


Yes, that would be a good thing. Such upgrade functionality should be
created in a modular fashion anyway so as to simplify adding to and
removing from it.


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Adrian Bunk

2004-04-01, 10:39 am

On Thu, Apr 01, 2004 at 08:03:02AM +0200, Adrian 'Dagurashibanipal' von Bidder wrote:
>...
> I think Debian should not consiously break upgrades over two major
> versions if it is so trivial to leave the support in.
>...


Debian does not support upgrades that skip one major major version.

E.g. last I checked, potato -> sarge breaks somewhere at the
app-defaults stuff in xlibs (an there might be many other breakages).

> cheers
> -- vbi


cu
Adrian

--

"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Adrian Bunk

2004-04-01, 11:35 am

On Thu, Apr 01, 2004 at 08:01:00AM +0200, Marc Haber wrote:
> On Wed, 31 Mar 2004 16:16:02 +0200, Morten Brix Pedersen
> <morten@mbrix.dk> wrote:
>
> I tend to keep these conflicts in to cater for the stupidity of our
> users. To give a reason: exim4's maintainer scripts create a user
> account. This account creation fails if the version of passwd is in a
> certain range due to a bug in passwd. The buggy versions are newer
> than woody, but older than testing, so any Debian system running a
> current version of any distribution is fine.
>
> However, there have been gazillions of bug reports against exim
> because people keep running outdated versions of unstable. We hat to
> add a versioned conflicts just to keep the number of bug reports down,
> while that conflict is technically not necessary any more. And no,
> that conflicts is bound to stay for a while. Won't be removed any time
> soon.


Conflicts for post-woody versions are definitely required for
Debian 3.1.

Due to stable being completely outdated, many people are in a
backport-hell, or they followed the message from some Debian developers
who told that stable + testing with apt pinning would solve their
problems.

> Greetings
> Marc


cu
Adrian

--

"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Colin Watson

2004-04-01, 11:35 am

On Thu, Apr 01, 2004 at 05:33:12PM +0200, Adrian Bunk wrote:
> On Thu, Apr 01, 2004 at 08:03:02AM +0200, Adrian 'Dagurashibanipal' von Bidder wrote:
>
> Debian does not support upgrades that skip one major major version.


That's no reason to make it intentionally worse, at least as far as I'm
concerned.

--
Colin Watson [cjwatson@flatline.org.uk]


--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Josip Rodin

2004-04-01, 4:39 pm

On Wed, Mar 31, 2004 at 04:16:02PM +0200, Morten Brix Pedersen wrote:
> Many old packages in Debian, have conflicts for specific version of
> packages that are pre-potato. Would it then be OK to submit a minor bug
> and tell the maintainer that the conflict could be removed?


No, because it's not a _minor bug_... a minor wishlist perhaps...
and a minor waste of time, the cynics would note.

--
2. That which causes joy or happiness.


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Adrian 'Dagurashibanipal' von Bidder

2004-04-02, 2:34 am

Branden Robinson

2004-04-02, 2:34 am

On Thu, Apr 01, 2004 at 05:33:12PM +0200, Adrian Bunk wrote:
> On Thu, Apr 01, 2004 at 08:03:02AM +0200, Adrian 'Dagurashibanipal' von Bidder wrote:
>
> Debian does not support upgrades that skip one major major version.
>
> E.g. last I checked, potato -> sarge breaks somewhere at the
> app-defaults stuff in xlibs (an there might be many other breakages).


The debian/TODO file in the xfree86 source package outlines a plan for
rectifying that.

--
G. Branden Robinson | Our ignorance is God; what we
Debian GNU/Linux | know is science.
branden@debian.org | -- Robert Green Ingersoll
http://people.debian.org/~branden/ |

Colin Watson

2004-04-02, 5:40 am

On Fri, Apr 02, 2004 at 08:33:01AM +0200, Adrian 'Dagurashibanipal' von Bidder wrote:
> [please don't cc: me. Isn't this the recommended practice on this list
> anyway?]


I wouldn't normally have done, but Adrian Bunk's Mail-Followup-To:
header included you.

--
Colin Watson [cjwatson@flatline.org.uk]


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Petter Reinholdtsen

2004-04-02, 6:34 am

[Adrian Bunk]
> E.g. last I checked, potato -> sarge breaks somewhere at the
> app-defaults stuff in xlibs (an there might be many other
> breakages).


Eh, upgrades to sarge do not necessarily work at the moment. I've
seen, reported and fixed upgrade problems when I have given it a try,
so I would not be surprised if there are still upgrade problems to be
fixed.

We need to do upgrade testing from Woody to Sarge before releasing
Sarge, and fix the issues we find.


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Adrian Bunk

2004-04-02, 10:38 am

On Fri, Apr 02, 2004 at 12:48:40PM +0200, Petter Reinholdtsen wrote:
> [Adrian Bunk]
>
> Eh, upgrades to sarge do not necessarily work at the moment. I've
> seen, reported and fixed upgrade problems when I have given it a try,
> so I would not be surprised if there are still upgrade problems to be
> fixed.
>
> We need to do upgrade testing from Woody to Sarge before releasing
> Sarge, and fix the issues we find.


That's clear.

I was talking about potato -> sarge, not woody -> sarge.


BTW:

One of the reasons why I dislike the _very_ short time between the
actual freeze and the release in the current release plan is that this
will give only very limited time of the comple set of packages that will
comprise Debian 3.1, including only few time to identify and fix all
upgrade problems.

It might happen that some problems start shortly before the freeze.

cu
Adrian

--

"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Wouter Verhelst

2004-04-02, 12:34 pm

Op vr 02-04-2004, om 17:15 schreef Adrian Bunk:
> I was talking about potato -> sarge, not woody -> sarge.


That is not supported, nor has it ever been. If you ever skipped
releases in an upgrade, and it worked, you've simply been lucky.

--
EARTH
smog | bricks
AIR -- mud -- FIRE
soda water | tequila
WATER
-- with thanks to fortune


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Adrian Bunk

2004-04-02, 1:38 pm

On Fri, Apr 02, 2004 at 06:46:00PM +0200, Wouter Verhelst wrote:
> Op vr 02-04-2004, om 17:15 schreef Adrian Bunk:
>
> That is not supported, nor has it ever been. If you ever skipped
> releases in an upgrade, and it worked, you've simply been lucky.


That's exactly what I said.

cu
Adrian

--

"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Martin Albert

2004-04-03, 10:33 am

> >For example, the debconf package has Conflicts: menu (<= 2.1.3-1).
>
> Looks perfectly useful to me. It is most definitely not a bug.


Especially if you install an 8MB machine from potato floppies in order
to upgrade then.

Have a nice day, martin


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Blars Blarson

2004-04-03, 3:34 pm

In article <20040331141602.GB25147@kolon.mbp> morten@mbrix.dk writes:
>Many old packages in Debian, have conflicts for specific version of
>packages that are pre-potato. Would it then be OK to submit a minor bug
>and tell the maintainer that the conflict could be removed?
>
>For example, the debconf package has Conflicts: menu (<= 2.1.3-1). But
>even potato has menu 2.1.5-10.1, so the conflict is not of much use
>anymore.


Instead of filing bugs (which should at most be wishlist), why don't
you submit a wishlist request to add a check to lintian and/or linda.
"Warning" is to strong for these, maybe they need a new level
"observation". The check is more likely to be added if you submit
patches.



--
Blars Blarson blarson@blars.org
http://www.blars.org/blars.html
With Microsoft, failure is not an option. It is a standard feature.


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Matt Zimmerman

2004-04-03, 4:36 pm

On Wed, Mar 31, 2004 at 06:35:41PM +0200, Martin Albert wrote:

>
> Especially if you install an 8MB machine from potato floppies in order
> to upgrade then.


This conflict is irrelevant in that scenario. You can only upgrade a potato
system to woody, and woody has the necessary conflict.

--
- mdz


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Jeroen van Wolffelaar

2004-04-03, 6:34 pm

On Sat, Apr 03, 2004 at 12:18:34PM -0800, Blars Blarson wrote:
> In article <20040331141602.GB25147@kolon.mbp> morten@mbrix.dk writes:
>
> Instead of filing bugs (which should at most be wishlist), why don't
> you submit a wishlist request to add a check to lintian and/or linda.
> "Warning" is to strong for these, maybe they need a new level
> "observation". The check is more likely to be added if you submit
> patches.


Which will be very hard, because it requires a full list of package
versions of at least potato, something the current lintian doesn't cater
for (and I'm not sure this will be accepted).

--Jeroen

--
Jeroen van Wolffelaar
Jeroen@wolffelaar.nl (also for Jabber & MSN; ICQ: 33944357)
http://Jeroen.A-Eskwadraat.nl


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Colin Watson

2004-04-04, 6:35 am

On Sat, Apr 03, 2004 at 12:18:34PM -0800, Blars Blarson wrote:
> Instead of filing bugs (which should at most be wishlist), why don't
> you submit a wishlist request to add a check to lintian and/or linda.
> "Warning" is to strong for these, maybe they need a new level
> "observation".


lintian has Informational (I tags, shown only with -I/--display-info.

--
Colin Watson [cjwatson@flatline.org.uk]


--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Sponsored Links






Free braindumps | Software forum | Database administration forum

Copyright 2003 - 2009 webservertalk.com